Many industrial processing plants utilize activated carbon to purify waste water by means of adsorption. However, activated carbon must be either replaced with new activated carbon or regenerated to remove adsorbed impurities and to replenish it's absorption properties. Since, activated carbon is costly, processors have been searching for cost effective methods for regenerating the spent activated carbon. Currently, regeneration of spent activated carbon requires that the activated carbon be heated to a temperature of from 400 to over 1000.degree. C. in the presence of steam.
The activation of carbon is an old process that has been carried out in many ways. To produce activated carbon, the carbon feedstock is a carbonaceous substance from which volatile matter has been substantially removed by heating in the absence of air (charring or carbonization), or from which volatile matter is naturally substantially absent. Examples of the former category are coconut char, wood char (i.e., charcoal), and bituminous coke. An example of the other category is anthracite coal of sufficiently low volatile content. The activation is effected by a gasification process that creates a greatly enlarged surface area and an extensive network of submicroscopic pores.
U.S. Pat. No. 4,192,962 (Nakao), filed Mar. 11, 1980, and U.S. Pat. No. 4,261,857 (Nakao), filed Apr. 14, 1981, describe a process where activated carbon is regenerated by passing the particles through a furnace equipped with electrodes that impart heat to regenerate the carbon. The electrodes are placed along the walls of the furnace and define one or more zones. The upper zone is expected to consume the most energy as regenerated activated carbon often contains significant moisture, which must be removed, and the carbon's temperature must be increased from roughly ambient to a target regeneration temperature between 500 to 1,000.degree. C. The Nakao patents describe a system for moving carbon through the furnace entirely by gravity and withdrawal through a single rotary valve. Nakao also describes a furnace where the inner cross-sectional area is gradually decreased and then gradually increased so as to form a throat. This causes a desirable temperature gradient to be formed within the furnace.
Suzuki et al. (U.S. Pat. No. 4,025,610; May 24, 1977) describes a vertical coking furnace consisting of a tube with opposing electrodes positioned at the top and bottom. Although this furnace was primarily operated for the denitrifying of metallurgical coke, it has features similar to an electric furnace applied for the production of activated carbon. Cordier et al. (U.S. Pat. No. 4,867,848; Sep. 19, 1989) also describe a vertical coking furnace with an upper portion used to preheat and devolitalize raw ovoids of coal. This is followed by a median section that is electrically heated and used for carbonizing and coking the ovoids, while sustaining a counter-current flow of recycled product gases recovered from the top of the furnace. A cooling chamber is located at the bottom of the furnace.
Gaylord et al. (U.S. Pat. No. 5,089,457; Feb. 15, 1992, and U.S. Pat. No. 5,173,921; Dec. 22, 1992) describe an apparatus and process for the production of activated carbon using an electrical resistance furnace. In this process, the carbon flows downward by gravity through a cylindrical furnace while an electric current is caused to pass through the bed of carbon particles. Steam is passed into the bottom of the furnace and passes upward through the furnace. Product activated carbon emerges from the bottom of the furnace. Gaylord specifies that the carbon feedstock must have been pre-processed to a temperature of from 550 to 750.degree. C. to achieve low electrical resistance and that the furnace inner radius be no more than 75 times the size-average of the largest dimensions of the carbon feedstock, wherein the size-average is defined as L(s)=SUM[L.sup.2 ]/SUM[L], wherein L is the largest dimension of a given particle. Assuming that the largest dimensions of the influent carbon particles are an average of 0.05", the furnace should be no greater than 3.75".
Gaylord also describes an apparatus consisting of a preheating chamber located above a reactor vessel supplied with steam at its bottom. The steam is allowed to rise through the reactor vessel and preheater. Electric current is passed between an electrode surrounding the preheating vessel and another located on the outside of the tower reactor vessel.
Gaylord's process and apparatus experience a variety of problems, including limited operating temperatures as a result of materials selection, uncontrolled fluidization of the activated carbon by the rising steam, non-uniform electric field gradients and carbon-flow patterns resulting in widely varying time-temperature histories for particles falling through the proposed reactor. In addition, by using a single electrical-resistance heating zone, Gaylord had limited ability to adjust for variations in energy consumption and temperature gradients along the vertical axis of the reactor.
Another related patent includes that of Hirai (U.S. Pat. No. 5,190,901; Mar. 2, 1993). Hirai describes a method of manufacturing activated carbon consisting of first carbonizing the raw material into an electrically conductive char, followed by activation within a batch furnace equipped with electrodes, and designed to allow the introduction of steam. Hirai (U.S. Pat. No. 5,287,383; Feb. 15, 1994) describes an apparatus for producing activated carbon or regenerating activated carbon. In one embodiment, the carbonizing furnace is a batch furnace followed by an activating unit fitted with a screw conveyor, electrodes, and steam supply. Alternatively, the activating unit is a tunnel furnace with a chain conveyor, electrodes, steam supply, and cooling chamber.
Hiralawa (U.S. Pat. No. 4,127,737; Nov. 28, 1978) describes a rotary drum reactivation furnace equipped with spaced helical electrodes fixed to the inner wall of the drum. Steam can be injected to promote reactivation, while the helical spacing of the electrodes served to propel the carbon from the inlet toward the outlet of the drum.
Du Plessis (U.S. Pat. No. 5,406,582; Apr. 11, 1995) describes an apparatus and process for the activation of carbon in a tubular furnace with two or more descending sections filled with carbon. Each section is equipped with inlets to independently introduce steam into each section and each section is fitted with a graphite electrode positioned at the center and top of each section and attached to a steep shelve equipped with openings to allow gases and vapors to escape.
Another singular apparatus is the Minfurn activated carbon regeneration furnace developed in South Africa and described in U.S. Pat. No. 5,317,592 of Van Staden (May 31, 1994). In this system, spent carbon is regenerated by passing the carbon by gravity between spaced electrodes in a tubular furnace. The control system utilizes the current flow as an indirect measure of the electrical resistance of the carbon, which obviates the need for direct temperature measurement within the furnace. Moss (U.S. Pat. No. 4,807,246; Feb. 21, 1989) also describes a South African furnace for treating granular activated carbon and consisting of a vertical column with a pair of spaced electrodes. Carbon flows through these electrodes by gravity and out a valve located at the bottom of the furnace. A central cone is provided to control the flow of carbon and to prevent serious segregation.
Baxter et al. (U.S. Pat. No. 5,579,334; Nov. 26, 1996) describe a rotary furnace equipped with spaced electrodes that allow direct resistive heating of a solid particulate medium. Hirakawa (U.S. Pat. No. 4,127,737; Nov. 28, 1978) also describes a rotary furnace, but with a series of helically spaced electrodes that are affixed to the inner wall of the furnace drum. The electrodes not only provide resistive heating, but also propel the material toward the outlet of the furnace. The furnace is designed to dry and reactivate activated carbon.
Quisser et at. (U.S. Pat. No. 4,655,968; Apr. 7, 1987 and U.S. Pat. No. 4,760,585; Jul. 26, 1988) describe a furnace equipped with electrodes for direct electric heating. However, this furnace is designed for the combustion of the incoming carbon and reduction of the residual to a slag for disposal.
A multi-electrode furnace is described by Eirich et al. (U.S. Pat. No. 4,624,003; Nov. 18, 1986). The furnace is equipped with generally planar pairs of electrodes that are mounted at an angle from the furnace walls. Each pair of electrodes is electrically isolated from the other electrode pairs. The electrical energy applied to each pair of electrodes can be separately adjusted. Din and Eirich (U.S. Pat. No. 5,694,413; Dec. 2, 1997) describe a furnace with at least one pair of electrodes arranged one above the other and are used to heat an electrically conductive medium passing by gravity through the furnace.
Mihara et al. (U.S. Pat. No. 4,398,295; Aug. 9, 1983) describe an apparatus for regenerating activated carbon and comprising a desorption tank having at its bottom a sloped tubular furnace lined with electrodes. As the activated carbon flows through the furnace tube, it is intercepted by transverse weirs at spaced intervals. Movement of the tank causes the activated carbon to flow along the furnace tube and causes the carbon particles that have been regenerated and have thereby become lighter, to flow over the weirs in preference to particles that are more dense and have not been completely reactivated.
Mizuno et al. (U.S. Pat. No. 4,139,489; Feb. 13, 1979, and U.S. Pat. No. 4,149,023; Apr. 10, 1979) describe a method of continuously regenerating activated carbon by passing particles through vertically spaced electrodes located in a tubular furnace. Particles flow through apertures of a defined size within the electrodes, filling the space between the electrodes.
Aubry et al. (U.S. Pat. No. 4,357,210; Nov. 2, 1982) describe a furnace for the calcination of carbonaceous materials and containing electrodes that pass current through the charge, while passing a non-reactive gas in the opposite direction to the movement of the charge.
Halm (U.S. Pat. No. 4,472,245; Sep. 18, 1984) describes a continuous process for the thermal treatment of carbonizable material. The process uses a vertical furnace and is charged with wood, cellulose, or some alternative substrate, The furnace has an upper and lower electrode. The charge moves slowly downward. Heat transfer to the upper portion of the furnace, causes the material to carbonize at the upper portion of the unit and become electrically conductive. This carbonized material eventually reaches the space between the electrodes where it is heated and exposed to steam. Movement through the furnace is induced by withdrawal of product from the bottom of the furnace.
It can be seen that direct resistive heating of carbon within a furnace is an established method in the art. However, it remains rarely applied in actual practice in comparison with more conventional rotary direct and indirect fired furnaces or multi-hearth furnaces. This failure to achieve significant commercial implementation, is the result of many deficiencies in the practical elements of electric furnaces. The current invention attempts to correct these deficiencies.
One of the first problems with electric furnaces operating with a flooded volume of particulate material relates to efficiently passing steam through the particulate mass. It is often observed that in many situations, steam moving in the up-flow direction emerges with great force from the top of the furnace carrying fluidized hot carbon particles. In practice, conventional electric furnaces, most of which have operated in the gas-phase upflow direction, have greatly reduced steam and gas throughput to avoid fluidization. Rotary and multi-hearth furnaces do not suffer from these limitations because the particulate mass does not occupy the entire volume of the furnace, allowing for steam to pass through a particle-free zone of the furnace.
Particle-flooded electric furnaces that heat the carbon using direct resistive heating have been generally confined to activated carbon regeneration, which requires far less steam and lower temperatures than activated carbon production. The problem becomes clear when one calculates the pressure drop required to pass sufficient steam through a bed of carbon particles. The present inventor has calculated the pressure drop using the Ergun and Orning equation (Ergun, S. "Fluid Blow Through Packed Columns", Chem. Eng. Progress, 48(2), February 1952), where: EQU dP/L=[150(1-f).sup.2 Gu]/[g.sub.c f.sup.3 P.sub.g (D.sub.p).sup.2
G specific mass flow rate in lbs./hr.--sq. ft. PA1 u=fluid viscosity in lbs./ft.--hr. PA1 g.sub.c =gravitational acceleration in ft/hr.sup.2 PA1 P.sub.g =fluid density in lb/ft.sup.3 and D.sub.p =particle diameter in ft.
TABLE I ______________________________________ Sample Furnace Conditions ______________________________________ Steam Flow 6,000,000 lbs. steam/gas at 1200.degree. K. Furnace Internal Diameter 2.00 ft. Steam Supply Pressure, 75 psi Specific Mass Flow Rate, G 300 lbs/hr - sq. ft. Void Fraction, f 0.40 Fluid Viscosity for Steam/Gas Mix, u 0.10 lbs./ft. - hr. Acceleration of Gravity, g.sub.c 4.15 .times. 10.sup.8 ft/hr.sup.2 Fluid Density, P.sub.g, STP 0.008 lbs. ft..sup.3 Particle Diameter, D.sub.p 0.04 ft. ______________________________________
Table 1 above set forth the conditions of an electric furnace having an internal diameter of 24 inches and producing 3,000,000 pounds of activated carbon per year. Using the conditions set forth in Table 1, one can solve the steam flow equations, by assuming that steam inlet pressure must be approximately equal to the pressure drop through the bed of carbon. Fluid density and viscosity of the elevated steam pressure must be corrected, which is simplified when the steam gauge pressure is set to equal dP and when one assumes that the change in fluid viscosity at these low pressures can be neglected.
This solution yields a steam gauge pressure of 4.0 psig/ft of reactor length under the Table 1 conditions. This large pressure drop makes it impractical to operate electric furnaces for carbon production. Not only is the volume of steam too large, but passing this steam in the up-flow direction greatly exceeds the fluidization limit. The furnace contents will become fluidized when dp/L is greater than bulk density/144. Since the bulk density of activated carbon is perhaps 30 pounds/ft.sup.3, the maximum upflow pressure drop is only 0.2 psid. The required steam flow is approximately twenty times greater than the fluidization limit.
One possible solution to the fluidization problem evident in the prior art is to operate the furnace at reduced steam flow, but this would directly reduce the capacity of the furnace to produce carbon. Alternatively, the furnace can be operated to regenerate carbon, which requires far less steam than production of new activated carbon from a suitable substrate. Another possibility is to break the furnace into several sectors with separate injection and removal of steam, each sector receiving only a fraction of the total steam required for the process. A combination of the two latter methods has often been used, but when producing activated carbon, rather than carrying out regeneration, productivity is certain to be greatly reduced by this steam problem.
Another possibility is to pass the steam in the down-flow direction, co-current with the flow of activated carbon granules. However, this will result in substantial dynamic forces upon the bed of particles. These forces accumulate in the direction of steam flow and must be dissipated by frictional forces between the carbon particles and the furnace walls, reactor internals, and supported by the carbon particles themselves. Unfortunately, activated carbon particles are not strong and serious crushing would be experienced if steam differential pressures are too large. For example, total dynamic dP at the bottom of a furnace will rise to 63.2 psid, and would be combined with a static pressure of 4.2 psi, if the carbon bed in Table 1 is 20 ft. in depth. Activated carbon particles are generally not able to withstand these crushing forces.
Steam flows of these proportions would also be expected to cause substantial attrition as a result of lateral von Karmen forces. These forces cause adjacent particles to oscillate and the resulting collisions would be expected to cause unacceptable formation of powdered activated carbon within the furnace. However, it has been found that von Karmen forces, and their resulting attrition, can be minimized by processing the carbon in a sufficiently short time frame, i.e., in a few hours. This leaves the problem of avoiding high dynamic crushing forces.
To minimize fluidization effects, it has been discovered by the present inventor that steam must be passed down through the carbon bed. To avoid a large accumulation of dynamic pressure at the bottom of the furnace, the dynamic force of the steam passing through the bed must be interrupted by internal structures that serve to dissipate the force at regular intervals. Because the carbon particles are not a conventional fluid, a force developed within the particle bed can be efficiently transferred from the particles to the walls of the furnace or to suitable baffles located at regular intervals along the length of the furnace. This transfer of downward force dissipates the dynamic pressure and prevents the development of pressures exceeding the crush strength of the carbon.
Whatever structures are used to dissipate dynamic loads, they must also allow the efficient movement of the carbon down the length of the furnace without causing serious non-uniform particulate flow patterns or wall effects. If possible, they should serve to dissipate these anticipated wall effects. The present inventor has discovered that a series of electrodes that simultaneously absorb this dynamic load, pass the carbon particles, and establish good electrical connections to the particle bed would satisfy the above structural requirements.
The present invention provides a method and apparatus which is capable of reactivating carbon with 50% more efficiency than conventional reactivation system. The present inventor has discovered that such an efficient system may be created by feeding steam co-currently with the downflowing carbon particles in the electric furnace so that no fluidization of the particles occurs. This is accomplished using a novel graphite electrode design that provides maximum contact with the downflowing carbon.
The present invention also provides many additional advantages which shall become apparent as described below.